Identifying Emotion by Keystroke Dynamics And Text Pattern Analysis

Abstract

Emotion is a cognitive process and is one of the important characteristics of human being that makes them different from machines. Traditionally, interaction between human and machines like computer do not exhibit any emotional exchanges. If we could develop an intelligent system which can interact with human involving emotions, that is, it can detect user emotions and change its behavior accordingly, then using machines could be more effective and friendly. Affective computing is the field that deals with this problem of identifying user emotion through various methods. Many steps have been taken to detect user emotions. Our approach in this paper is to detect user emotions through analyzing the keystroke patterns of the user and the type of texts (words, sentences) used by them. This combined analysis gives us a promising result showing substantial number of emotional states detected from user input. Several Machine learning algorithms of Weka were used to analyze keystroke features and text pattern analysis. We have chosen keystroke before it is the cheapest medium of communication with computer. We have considered 7 emotional classes. For text pattern analysis we have used vector space model (VSM) with jaccard similarity. Our combined approach showed above 80% accuracies in identifying emotions.